AI Project Management Agent (Digital Project Coordinator) - Full-Stack + AI Build

We are building an AI agent that acts as a digital project coordinator for complex, multi-functional product development programs. This is not another task tracker. The agent should generate plans from a short description, turn meetings and emails into assigned tasks, track milestones and stage-gates, flag risks early, forecast launch dates, and auto-generate executive reports. We have a complete project plan and data model already prepared (24-week roadmap, sprint breakdown, architecture). We will share the full documents in messages once we begin discussing scope. WHAT WE NEED We are looking for a developer (or a developer + designer pair) to build production v1. Core capabilities: - Conversational agent that generates a gated project plan (phases, gates, owners, dependencies) from a paragraph - Ingestion + extraction pipeline: email and meeting transcripts converted into structured, assigned tasks with source citations - Stage-gate readiness tracking (item-by-item, with evidence linking) - Risk register, critical-path detection, and launch-date forecasting - Manager + executive dashboards, portfolio rollup, and automated status reports - Proactive follow-ups and escalations RECOMMENDED STACK - Frontend: Next.js (React) - Backend: Node.js (NestJS or Express), TypeScript - Data: PostgreSQL + object storage + a vector store (pgvector or managed) - AI: LLM API with tool/function calling + embeddings for RAG - Integrations: Microsoft Graph (Outlook mail, Teams, calendar), file repository WHAT TO INCLUDE IN YOUR PROPOSAL 1. A fixed-price quotation (or hourly rate + estimated hours) for production v1 2. Your estimated timeline / availability 3. Relevant experience: LLM agents, tool-calling, RAG pipelines, and Microsoft Graph integrations (please link 1-2 examples) 4. Whether you work solo or with a designer 5. Any questions on scope We have detailed checklists, a playbook, and the full project plan ready to share in messages. Please confirm you have direct experience building LLM-powered agents (not just API wrappers). Looking forward to your quotation.